International Law

The US Government Position on Imminence and Active Self-Defense

By Kenneth Anderson
Thursday, February 7, 2013, 12:26 PM

(Updated and extended.)  The White Paper's reference to imminence has occasioned some heated rhetoric about the Obama administration stretching the notion beyond all possible ordinary meaning or bounds, etc.  But it's worth bearing in mind that there's nothing new in this from the standpoint of the US government.  The US government has held this view since the early 1980s at a minimum, and has articulated it in official speeches, including the famous 1989 address by then-DOS legal adviser Abraham Sofaer, as well as in an earlier speech (the 1984 Park Avenue Synagogue address) by then-Secretary of State George Shultz.  (Thanks to Greg McNeal for digging out the language; Bobby also makes reference to this in his Awlaki article.)  As Sofaer said clear back in 1989 (in the same interagency-cleared speech that stated the unable-or-unwilling standard):

Some States, such as Lebanon, are simply unable to exercise authority over terrorists, even if they were inclined to do so. The United States must be free to utilize force with sufficient flexibility to defend itself and its allies effectively against threats resulting from such breaches of international responsibility. As Secretary of State George P. Shultz predicted in 1984: “We can expect more terrorism directed at our strategic interests around the world in the years ahead. To combat it, we must be willing to use military force.” ....

In 1984 Secretary Shultz described this policy as an “active defense.” “From a practical standpoint,” he said, “a purely passive defense does not provide enough of a deterrent to terrorism and the states that sponsor it.” Later that year he described why an active defense was needed to deter:

"We must reach a consensus in this country that our responses should go beyond passive defense to consider means of active prevention, preemption, and retaliation. Our goal must be to prevent and deter future terrorist acts, and experience has taught us over the years that one of the best deterrents to terrorism is the certainty that swift and sure measures will be taken against those who engage in it. We should take steps toward carrying out such measures. There should be no moral confusion on this issue. Our aim is not to seek revenge but to put an end to violent attacks against innocent people, to make the world a safer place to live for all of us. Clearly the democracies have a moral right, indeed a duty, to defend themselves."

Whatever one thinks of this standard - it seems to me correct - it is not a new position for the United States government.  Neither the Obama administration nor the Bush administration before it has created a new doctrine or US interpretation of international law.  (Added:  Kevin Jon Heller points me to a post by Marko Milanovic at EJILTalk, pointing out, among other things, that in the case of state consent, imminence does not come into play.  It's a useful read and I recommend it; while I'm at it, I commend EJILTalk to you as a terrific source of international law analysis on many different topics.)


But why embrace this notion of "active self-defense" at all?  What is it supposed to do for a state in practical ways?  In practical and strategic (not legal) terms, it does not allow the terrorist group what amounts to a monopoly on strategic initiative.  This is fundamental the point Shultz makes.

We can add that there is a spatial as well as temporal aspect to this.  Waiting to act until the last possible moment allows the terrorist enemy to pick where it strikes and when, and forces the dispersal of the defender's forces to try and defend against many possible targets.  Acting not just against a particular threatened attack, but against the group as suchits organization, leadership, safe havens, and so on, allows the ability to disrupt the group's ability to attack at all or even to destroy it functionally.  It's not just the distinction between operations on offense versus defense; it's the difference between acting strategically and acting merely tactically.  Moreover, active self-defense implies an institutional rather than merely transactional approach to counterterrorism.  (I raised a number of these questions as a matter of risk analysis in an old essay challenging the idea of a narrow, defensive, "serial catastrophism," tactical cost-benefit approach to counterterrorism; it didn't receive much attention then and had muddles of its own, but the basic ideas perhaps deserve revisiting.)

All this is well known, of course, because it defines the US's "counterterrorism-on-offense" strategy.  It has been a successful strategy, in my view, in part because, for reasons I discuss in this short Hoover Institution essay as well as this Hoover podcast, it is far from being a drones-alone strategy.  Drone warfare has proved crucial and will continue to be so.  But that strategy also involves crucial, long-term strategic elements of "territorial denial" to terrorist groups seeking haven - in the form of military advisers and assistance to governments battling insurgents who also either have transnational terrorist ambitions or might harbor those who do; intelligence gathering and the establishment of intelligence networks on the ground in areas of emerging threats; proxy forces; etc.  Note that a fundamental goal of all this is to avoid further US counterinsurgency wars as a strategy of counterterrorism.  The difficulty for the US strategy today is that although the advantages of drone warfare have become clear, there is still an irreducible element of "territorial denial" to terrorist groups, and if the US will not do it directly with boots-on-the-ground warfare, it still has to fill that gap.  The cutting edge strategic questions and, if commentators in the US were more aware of it, the cutting edge legal policy questions actually run far more to these other activities than they do to drone warfare.

But the legal framing matters with respect to targeting of all kinds.  Overall, the imminence requirement - and hence a restriction on the US's ability to strike strategically and not merely in tactical, seriatim, prevent-the-next-attack fashion - does not apply if there is an armed conflict already underway.  The criteria of imminence, broad or narrow, are not at issue, because the armed conflict is already engaged - whatever question of imminence arose, that moment is long since passed.  That has been the general view of the Obama administration in its pursuit of counterterrorism-on-offense; the armed conflict is underway and it's game-on - it's not game-on, game-off, game-on.  It is defined as a unitary conflict in space and across time under the AUMF.  Many people are skeptical of that view, of course.  But Kevin Jon Heller is quite right in his observation at Opinio Juris that, if the US government actually means its invocation of a non-international armed conflict (a position Kevin does not share, but takes by assumption arguendo), the White Paper's discussion of imminence and capture is not relevant or is at least confused, at least as a matter of international law.


The Obama administration thus hiccups at particular moments, and resurrects imminence within the context of an armed conflict in which one would have thought imminence already answered.  After all, it's already game-on, from the US government's overall legal standpoint.  The US government peculiarly seems to invoke imminence in certain circumstances running to individual persons and the particular plots and other things they might be threatening - "individual" and "transactional" view of imminence within the existing armed conflict.  It is hard to make sense of this on law of war grounds.  As an international law matter, as Kevin says, imminence has salience on the front end, before an armed conflict has begun or as the opening of an armed conflict and its jus ad bellum considerations, meaning whether it is lawful to resort to force.  With that as the starting point, the main observation of this post is  that when imminence is a relevant legal question, the US has long held a capacious "active self-defense" view of its meaning and interpretation.  But the requirement of imminence is not usually thought to arise in the law of war once the conflict is underway.

The administration seems to invoke imminence in two distinct settings within the current armed conflict.  One is the case of an American citizen, in which it is offered (according to the White Paper and (I think a fair reading) also Holder's Northwestern speech) as a further US constitutional law hurdle upon otherwise lawful targeting in armed conflict.  Greg McNeal notes that the White Paper explicitly treats imminence as a domestic law requirement - part of its due process analysis (see Trevor Morrison's post here at Lawfare).  In that case, it is not at odds with the usual reading of international law on imminence; it is a different usage and legal paradigm laid atop of the armed conflict targeting regime.  However, much of the language of the White Paper is unclear as to whether it is strictly domestic law analysis about domestic law carried abroad with regards to a US citizen, or whether it is attempting to make an analogy from domestic law about international law of imminence; either way, it seems to me unpersuasive.  Rick Pildes (see his Lawfare post) seems to me quite right in his skepticism about this kind of analogy, which in effect draws from domestic law and institutions and then is sent abroad; the kind of due process analysis upon which these domestic law settings require, at a very minimum, sovereign territorial control.  The kinds of lines drawn and balancing tests of domestic rule of law analysis take as their prior assumption that the writ of the rule of law runs across sovereign territory; one cannot simply assume them into places where that sovereign territorial control is lacking.

On the other hand, I cannot agree with John Yoo's WSJ essay declaring that due process requires judicial participation; Eric Holder was right to say that not all due process is judicial process.  John holds out for sharply alternative legal paradigms here.  You either choose armed conflict, with its targeting rules, or else you choose some form of domestic criminal law, with its insistence on judicial reviews to satisfy due process.  What you cannot do, he avers, is construct an ad hoc middle ground - in which it's not okay simply to target according to conventional war rules, including American citizens, but on the other hand, these American citizens and potentially other non-citizen adversaries have due process rights, except that they don't involve the judiciary and lie entirely in the hands of the executive.  John's is a principled position, to be sure - go with one or the other (though it should be clear that for him this is armed conflict), but don't mix and match according to your policy and political preferences to get the hand that suits you the most.

I would reject the premise that due process is necessarily judicial process.  But at a more general level, I don't think that the facts of US counterterrorism-on-offense, counterterrorism conducted abroad and using force, JSOC and CIA special operators, drone warfare, and so on, really lead to so clear a position either way.  This is not criminal law enforcement - quite so.  But likewise it's not conventional war in the most contentious parts, either, drone warfare and targeted killing.  It is a different form of the use of force, one that I have sometimes called "intelligence driven uses of force," or in colloquial terms, covert actions that depend upon intelligence, precision, discrete uses of force in the first place.  In these discrete, quasi-covert uses of force, I think it's perfectly okay to say, within the executive branch, we have to create a process by which to evaluate who is actually a lawful target in this armed conflict, and it's more difficult than a conventional war where the enemy is more easily identified.

That's targeting within a particular strategic and technological framework - the consequences of the rise of precision - and what arises from this ought to be an attempt to take seriously the international law principle of distinction and discrimination in this special kind of use of force.  I think that this kind of use of force implies a standard of discrimination in both selection of the target and engagement in attacking the target that ought to be higher than it can or should be in conventional war against anonymous cohorts of hostile forces.  This does not suggest that conventional war should ever be bound to this kind of essentially "individuated" target determination; nor is this individuated targeting implying that targeting is no longer status-based.  It merely recognizes that where one is selecting an individual target with a precision weapon, one has to have, to fulfill the principle of discrimination, some reason to select this target and not that one.  This has nothing to do with due process, a shift from status-based to conduct -based targeting; it is merely that selection requires a reason for selecting.  This is entirely a claim within the general law of war - an attempt to adapt conventional war targeting law to new possibilities of precision and new strategic reach with precision weapons that, moreover, have the ability to "persist" against objectives.


Put another way, the law of war needs to evolve a sub-branch, as ever based around the immutable principles of necessity, distinction and discrimination, proportionality, and humanity, but adapted to covert, discrete, and intelligence-driven uses of force.  To be crystal clear, this is not an attempt to do an end-run around the constraints of the laws of war  - it is, rather, an attempt to extend rules and principles to new technologies and strategic uses, as the law of war has long done.  In this case, its main burden is to raise the veil on the hitherto unregulated - because essentially invisible - conduct of covert action by intelligence services and bring regulation to it.  It is a step forward to recognize that such uses of force exist - and to say, as the Obama administration has been correct to do, that these are lawful uses of force as a category, but they must meet a minimum level of conduct, set by the law of armed conflict and its fundamental principles.  I believe that if John thought about it in these terms, rather than as something it is not - not really - viz., an attempt to shoehorn bits and pieces of the criminal law and domestic due process into the laws of war, he might agree that this is a salutary project in the laws of war that has nothing to do with domestic due process.

Due process becomes an issue, however, where a US citizen is the possible object of attack.  In conventional war, it has not been regarded as an especially difficult issue and ought not to be now (e.g., the case of a dual national who has taken up arms on the Taliban side in Afghanistan, shooting at US forces).  In these intelligence-driven, targeted uses of force, however, it does seem to me more than plausible to say that a US citizen abroad is entitled to some form of due process to establish the threshold matter that he or she is a lawful target - some form of due process running to the question of whether he or she is someone who can be treated as a lawful target under the laws of war.  I think that is a consequence of US citizenship, but I don't think it is a judicial question in the context of counterterrorism abroad.  The proper due process ought indeed to be a formalized one, but strictly between the two political branches, ideally built by statute into the oversight process.  Eric Holder was right to insist that targeting is almost by definition a core executive function, and the judiciary does not have a role in the name of due process.  (See Bobby's post on proposals for judicial supervision of targeted killing.  One additional question I would be grateful to see addressed is more basic than the ones Bobby takes up - viz., is there any reason to think that, particularly an ex ante, judicial warrant or judicial process for targeted killing violates Constitution separation of powers, even if the role is mandated by Congress, because it somehow impermissibly inserts the judiciary into something that is Constitutionally committed to the executive?)  But this really needs to happen, as Jack noted in the Washington Post yesterday, as part of a statutory overhaul of the oversight process.  It desperately needs updating and expanding to take account of a future that will feature many more instances of special operations - not just targeted killings with drones, but many other kinds of operations as the US leans more heavily on special operations in an effort to stay out of more counterinsurgency ground wars.

Note, however, I agree with John Yoo that this is not a matter of criminal law due process with regard to whatever due process is owed to a US citizen who has fled abroad (as with Awlaki, and who is engaged in senior operational roles - such as meetings with the Detroit underwear bomber, among other things).  It's not a matter of judicial due process, and we should not make an attempt to graft domestic criminal law onto targeting in war.  Targeting is targeting; it is not a method of pursuing a law enforcement manhunt using, however, a super-super-highly armed posse, whose basic weapons are all those of warfare - but we then somehow tell ourselves that the operation still contemplates an arrest.  It's a category mistake and an exercise in political self-deception.  In that regard, it seems to me a mistake for the Obama administration to say that there is some special obligation to consider feasibility of capture of an American citizen, as though this were just some specially militarized posse out hunting a fugitive.  In that regard, the better approach is to accept that due process runs only as far as the determination of whether one is lawfully subject to lethal force, the same lethal force as other lawful targets.


Once that determination has been made, due process has been satisfied, and the relevant legal regime is the law of targeting. It's not a ramped up form of manhunt in a criminal law sense that will however,  almost certainly result in a kill.  Indeed, it is corrupting to the legitimacy of the process to suggest that there is an obligation to pursue capture over kill, and it is morally wrong take any great risk with US forces lives to try and make that a reality, when in fact it rarely could or would be - and that because we quite correctly do not believe there is an obligation, either legal or moral, to subject our forces to risks that reflecting levels of due process that implicitly assume complete sovereign control of territory - something the US possesses in fact and in law over its own territory, but not in Yemen or Somalia.  Better to recognize that due process is satisfied by a serious political oversight process and a special, formal review for American citizens to determine whether someone is a lawful target or not.

One can tweak this with special possibilities of safe surrender to authorities, or measures where feasible and consistent with security concerns to put an American citizen on notice that he or she might be subject to attack and ought therefore to surrender.  But notice will not always be feasible with someone who is in hiding abroad, for obvious reasons about informing the person to be extra cautious - putting them on notice of pursuit even while, for example, he strategizes operations in which to hide explosives in a jihadist's underwear on a plane or load explosives into a printer cartridge in airline cargo.

That's with regard to imminence raised as an issue for US citizens as a function of domestic due process. But the Obama administration, at least in some of its rumored interior discussions, seems to have raised imminence in at least one international law context with regards to individuals in what is otherwise seen as an armed conflict.  I refer to press reports of debates between Harold Koh and Jeh Johnson, or anyway their offices, over whether proposed strikes against foot soldiers (as they were sometimes described in the press) in Yemen had to be evaluated according to some standard of imminence or threat posed to the United States (and not merely to Yemen).  If, however, these individuals - even if individually threats merely against the Yemen government and not directly the US - are members or taking direct part in hostilities as part of "associated forces" under the AUMF, they are part of the conflict with the US and can be targeted as such; imminence is legally neither here nor there.

(There might be a special way in which an assessment of imminence might be thought to come into play, however.  Imminence in this context has passed from being a question of the conflict itself to the level of individuals as a matter of targeting and their individual "transactions," as I've called it. The standard for targeting within an existing armed conflict is not imminence as such.  However, if one holds a view of DPH that permits a person to rotate in and out of being a lawful target on some "part-time fighter" view then the imminence of a person's hostile acts should be regarded as a sort of temporal evidentiary factor in establishing a person's targetability status - to show, in effect, when targetable.  A part-time fighter is targetable when, other things equal, the threat he or she poses becomes imminent - and thus constitutes taking direct part in hostilities.  However, to be clear, the US does not hold to this kind of part-time fighter view of DPH; at the individual level, there's no rotating in-out-in-out.  The US has been clear on this since the negotiations over Additional Protocol I at least.)

Broadly and loosely speaking, then, the alternatives for being targetable are membership in a terrorist group (evidenced in some fashion; this runs beyond armed conflict law to legally distinct questions of terrorist groups); status-based membership in "hostile forces" in an armed conflict (with status evidenced in some fashion) that can be attacked as groups as such; conduct-based direct participation in hostilities by civilians (with further debates over whether you can be targeted when not directly participating, so-called "continuous combat function," and a variety of other questions raised in the preceding paragraph over part-time fighters).  As Greg McNeal has pointed out (and we await his paper on these matters), the US government uses a membership-based standard in some circumstances, and conduct-based standards in others - but so far as I am aware at this moment, imminence within the context of the existing AUMF armed conflict has been raised as a consideration in establishing targetability only with regards to US citizens (I am prepared to be corrected on this) and (though apparently settled to be a matter of policy and not a requirement of law) the policy of assessing the imminent threat of lower level fighters in Yemen (or potentially other places).


Imminence is about the assessment of a threat that one might conclude should be addressed with force.  It's before passing the point of "game-on." "Active self-defense," however, is nothing novel, and simply means that the estimation of imminence is able to take the past history of a group into account as well as what is known about it as a group, and not merely what is known about any particular plot or threat. It permits a threatened state to pick its own moments for striking, when and where, against not just a particular plot or threat (or a series of them, treated as separate individual "transactions," as it were) but against the actor itself.  It treats particular threats from an actor as linked and on-going, regards lawful responses in that way, and so does not force the state into a purely, impossibly reactive position.  It is not the governing doctrine in the case of an armed conflict underway, because if one is in an armed conflict, the defeat of the enemy as such - as the enemy and not merely as a series of individual threat transactions - is lawful and legitimate; war aimed at the defeat of the enemy as such embodies "active self-defense."

But, as I've pointed out in other contexts, there will be future threats to national security in which the President might determine to act in the absence of an already on-going armed conflict (what I have sometimes called "naked" self-defense, though not the aspect of self-defense that "naked" self-defense has attracted some criticism on different grounds than these; I'll post about that another day), against some wholly new terrorist group, for example, entirely uncovered by the AUMF or any existing state of armed conflict.  The President's legal authority to respond and to address the threat and, on some judgments of the nature of certain threats, address the threat by addressing the actor - at least so far as the US government is concerned - embraces this notion of holistic, active self-defense in the framing of the meaning of "imminence."